🎉 Please If you have any questions, visit the Patreon link provided in the Description section. You can join the chat group and always get your queries answered without any cost.
thanks so much, it works like a charm! I can't see any links to the Patreon like you say. and I have questions I would like to you to answer if possible. so please provide it. thanks.
The way you teach is incredible, I love the way you teach, thank you very much for the videos, your tutorials are certainly the best as they teach you how to do the workflows and it's not like you have to watch 30 minutes to buy a ready-made workflow
WOW. I was an idiot and had the UNET loader with the wrong model. THIS WORKS! AMAZING. Subbed and now checking out the rest if your tutorials. What everyone says below is true. Thank you!
I just got to say WOW! Great job on the video! Just the right speed, and you are one of the only ones who actually took the time to list links on where you can download everything needed to create this workflow. For a new venture we are starting up soon, this will come in handy! I will check more vids out tomorrow.
@@CgTopTips Welcome! I just got around to creating this work flow, and unfortunately, I am having some issues. Some of the nodes I didn't have already (completely new to comfyui), and I was able to get the ones missing installed. Unfortunately, 1 of the nodes that I didn't have (DualClipLoader (GGUF) was one of them. I installed it, but for some reason the only options it provides me for clip name 1 and clip name 2 is "null". I can most likely Google the file names I seen in your video, but have no clue what folder they need to be downloaded into it. Also the option you selected for the "Load VAE", I don't have that one. I did click the links below the video, in the description, but for some reason the file names weren't the same as you have them for the 2 "LoraLoaderModelOnly" I can create a small, quick video to show exactly what I am referring to, if that is alright with you. Just let me know the best method to send it to you.
@@CgTopTips So I was able to figure out how to get most of the installed and figured out. I realized for the "LoraLoaderModelOnly" that you renamed the file to Catviton. When I click the link to download the Flux1 Turbo Alpha, it has the same file name the the CatViton had. So I renamed it to the Flux 1 Turbo like I seen in the video. Not sure if this is causion me an isue or not, but I get en error message in the backend, and I am getting stuck on the Ksampler. It gets to that point and essentially wont move past it. Also when I found the filfor the "LoadVAE", it was named Flux1\ae so I left it named that.
@@CgTopTips Apologies in advance bro for all the comments lol. Just wanted to say, that some how, it ended up working. Just took longer than I was anticipating even though I got a pretty good gaming desktop pc. Except for not knowing if I did it correctly for the second LoraLoaderModelOnly, the rest of it is good to go. I guess you sped up the video for the generation lol. It is incredible on how the final image turned out. You would never know that she wasn't originally wearing the clothes!
I downloaded every thing and tried it , it wasn’t easy like the way it’s showing here ! I get extra hand or half cloths and the other half is not showing like cloudy . I have no idea what is the problem
thanks for the nice tutorial, this generally worked fine, however i have a vertical photo and top and bottom parts are getting cropped, how can i prevent that?
Well done, but what about output resolution and also if I'm using an image of a person with a tall image format the output gets cropped so the top and bottom is cut away 🤔
Irmão, estou voltando aqui após ter assistido esse vídeo há alguns dias, para te agradecer por esse conteúdo. Como é esclarecedor a forma como você apresenta, passo a passo, a construção dos nodes. Muito obrigado.
9:45 The "pandas" text on the top of the shirt was changed to "pawmas" and the "pandas" text on the bottom was changed to small, unintelligible letters.
Hello. Thanks for sharing. How do your runs give such fast results? Are you using Comfyui on your local computer or with Google Colab? If you are using Comfyui on your local computer, can you please provide your system information?
Hi, have you already made a video for face consistency? I'm trying to create the same character in multiple photos and positions. Thanks for your reply :)
In my experience, redux isn't necessary, and instead use the prompt that was trained by the creator of the lora. "The pair of images highlights a clothing and its styling on a model, high resolution, 4K, 8K; [IMAGE1] Detailed product shot of a clothing [IMAGE2] The same cloth is worn by a model in a lifestyle setting." Also, you might get slightly improved results if you disable noise_mask.
@@CgTopTipsI downloaded it and opened your workflow in seaart ai, it says "When loading the graph, the following node types were not found: StyleModelApplySimple / AddMaskForICLora" I don't know how to install comfyui in mac, so it is more complicated for me
Ensure the models you select closely match those in the video. Note: If you have any questions, visit the Patreon link provided in the Description section. You can join the chat group and always get your queries answered without any cost.
Less is more. Adding more and more controls for the AI to follow removes it's creativity and ofter lowers the output quality. Zooming in on the output of this workflow reveals the output quality is already quite low
@@kallamamran lot of people try to get consistency on the character, body, face, clothes and you can't achieve that with "creativity" so it all depend what you are trying to achieve.
When using it, I found it difficult to reproduce specific body shapes, such as large breasts and buttocks; after the change of clothes, the body shape tends to be more like an average figure.
@@CgTopTips hey I got it working and is working perfectly but I do have a question can I change the clothes as well in the video ? Is there a workflow for that too ?
🎉 Please If you have any questions, visit the Patreon link provided in the Description section. You can join the chat group and always get your queries answered without any cost.
thanks so much, it works like a charm!
I can't see any links to the Patreon like you say. and I have questions I would like to you to answer if possible. so please provide it. thanks.
The way you teach is incredible, I love the way you teach, thank you very much for the videos, your tutorials are certainly the best as they teach you how to do the workflows and it's not like you have to watch 30 minutes to buy a ready-made workflow
WOW. I was an idiot and had the UNET loader with the wrong model. THIS WORKS! AMAZING. Subbed and now checking out the rest if your tutorials. What everyone says below is true. Thank you!
Thank you very much! I've been searching for a related one for days now, and this one of yours is superior to other bloggers'.👍
I just got to say WOW! Great job on the video! Just the right speed, and you are one of the only ones who actually took the time to list links on where you can download everything needed to create this workflow. For a new venture we are starting up soon, this will come in handy! I will check more vids out tomorrow.
Thanks Bro
@@CgTopTips Welcome! I just got around to creating this work flow, and unfortunately, I am having some issues. Some of the nodes I didn't have already (completely new to comfyui), and I was able to get the ones missing installed. Unfortunately, 1 of the nodes that I didn't have (DualClipLoader (GGUF) was one of them. I installed it, but for some reason the only options it provides me for clip name 1 and clip name 2 is "null". I can most likely Google the file names I seen in your video, but have no clue what folder they need to be downloaded into it.
Also the option you selected for the "Load VAE", I don't have that one.
I did click the links below the video, in the description, but for some reason the file names weren't the same as you have them for the 2 "LoraLoaderModelOnly"
I can create a small, quick video to show exactly what I am referring to, if that is alright with you. Just let me know the best method to send it to you.
@@CgTopTips So I was able to figure out how to get most of the installed and figured out. I realized for the "LoraLoaderModelOnly" that you renamed the file to Catviton. When I click the link to download the Flux1 Turbo Alpha, it has the same file name the the CatViton had. So I renamed it to the Flux 1 Turbo like I seen in the video. Not sure if this is causion me an isue or not, but I get en error message in the backend, and I am getting stuck on the Ksampler. It gets to that point and essentially wont move past it. Also when I found the filfor the "LoadVAE", it was named Flux1\ae so I left it named that.
@@CgTopTips Apologies in advance bro for all the comments lol. Just wanted to say, that some how, it ended up working. Just took longer than I was anticipating even though I got a pretty good gaming desktop pc. Except for not knowing if I did it correctly for the second LoraLoaderModelOnly, the rest of it is good to go. I guess you sped up the video for the generation lol.
It is incredible on how the final image turned out. You would never know that she wasn't originally wearing the clothes!
Absolutamente increíble, eres mi canal favorito! Qué bueno, me quedo sin palabras
Once again reading my mind! Thank you so much!
🙏
Great work, thank you.
how about using SAM model or a finetuned one to automatically segment the body parts that needs to be repainted
I downloaded every thing and tried it , it wasn’t easy like the way it’s showing here ! I get extra hand or half cloths and the other half is not showing like cloudy . I have no idea what is the problem
That's amazing and looks perfect.
I've been using flux and catvton in a much more complicated way, but using it this way seems to work better for me.
thanks for the nice tutorial, this generally worked fine, however i have a vertical photo and top and bottom parts are getting cropped, how can i prevent that?
This is amazing! Thank you very much for tutorial!)
This works great, thanks so much!
this is the best virtual try-on workflow by far. I am wondering how to adjust the output photo resolution?
Well done, but what about output resolution and also if I'm using an image of a person with a tall image format the output gets cropped so the top and bottom is cut away 🤔
Irmão, estou voltando aqui após ter assistido esse vídeo há alguns dias, para te agradecer por esse conteúdo. Como é esclarecedor a forma como você apresenta, passo a passo, a construção dos nodes. Muito obrigado.
it works well ! big thanks !
thanks for your great tutorials, please can you share the workflow?
BRAVO!!!!!!!!!!
9:45 The "pandas" text on the top of the shirt was changed to "pawmas" and the "pandas" text on the bottom was changed to small, unintelligible letters.
You're right-regenerate until you achieve the desired result
Hello. Thanks for sharing. How do your runs give such fast results? Are you using Comfyui on your local computer or with Google Colab? If you are using Comfyui on your local computer, can you please provide your system information?
Thank you.
Is there anyway to retain the original resolution?
The photo was successful, but the photo result is blurry, can it be fixed?
I like the simplicity of this workflow, but there is a flaw, I have been over this a few times and the crop is not cropping to original size
Hi, have you already made a video for face consistency? I'm trying to create the same character in multiple photos and positions. Thanks for your reply :)
WOW Great Video, Workflow !!!
Please if you have any question regarding this video, let me know on Patreon chat group, It's free :)
I am using the workflow with the original dev fp8 model but I usually get a clothing with half of which is complete white color, what's the problem?
works very good! thanks :)
Hi I'm getting and error on command prompt about lora diffusion model.img in weight shape ( 3072 ,384 ) how do I solve that
Hello, thank you very much. How to increase the size of the image? The size is always 768x1024.
Where can I Download workflow? Your link is empty
Doesn't work or lower garments on HF space. Is it true for this workflow as well?
how do I get this new Mask Editor interface?
can u please make your workflows compatible with openart cloud? thanks alot.
Great video
In my experience, redux isn't necessary, and instead use the prompt that was trained by the creator of the lora.
"The pair of images highlights a clothing and its styling on a model, high resolution, 4K, 8K; [IMAGE1] Detailed product shot of a clothing [IMAGE2] The same cloth is worn by a model in a lifestyle setting."
Also, you might get slightly improved results if you disable noise_mask.
how to find location file
Will it work with GTX 1060 6GB also? Please reply 🙏🏻
I can't find "Add Mask for IC Lora node" option in Seaart Ai workflow, what is a similar option?
Did you download my workflow, or did you recreate it from scratch?
@@CgTopTipsI downloaded it and opened your workflow in seaart ai, it says "When loading the graph, the following node types were not found: StyleModelApplySimple / AddMaskForICLora" I don't know how to install comfyui in mac, so it is more complicated for me
@@knijaifajhj9u9uad9 It gives import failed error. The node is being loaded but not used. I click on Try fix but it doesn't fix it.
mat1 and mat2 shapes cannot be multiplied (257x1664 and 1152x12288) how to solve?
Ensure the models you select closely match those in the video.
Note: If you have any questions, visit the Patreon link provided in the Description section. You can join the chat group and always get your queries answered without any cost.
AMAZING WOW!!!!
Try to make it FLUX Fill + Redux + ControlNet (depth & canny) + SAM + LoRa 🤔
Less is more. Adding more and more controls for the AI to follow removes it's creativity and ofter lowers the output quality. Zooming in on the output of this workflow reveals the output quality is already quite low
@@kallamamran lot of people try to get consistency on the character, body, face, clothes and you can't achieve that with "creativity" so it all depend what you are trying to achieve.
When using it, I found it difficult to reproduce specific body shapes, such as large breasts and buttocks; after the change of clothes, the body shape tends to be more like an average figure.
Make a video about face swap
Did you watch this video ?
th-cam.com/video/3mfwpGPiQN0/w-d-xo.html
@CgTopTips ohh thanks I will watch it now
@@CgTopTips hey I tried it but can't download reactor node ? Showing error
Insightface or wheel error
@@CgTopTips hey I got it working and is working perfectly but I do have a question can I change the clothes as well in the video ? Is there a workflow for that too ?
if it works for video would be fun.
a few dollars to pay for your skill is not a problem. but most posts are 1 year old and depricated. make a current and tested post, please
Sure Bro
th-cam.com/video/a2dHkg492lI/w-d-xo.html